From f95bf754302ceaf79a5d379e8d357d3103b188d2 Mon Sep 17 00:00:00 2001
From: Huan Chen <huanc@google.com>
Date: Mon, 7 Oct 2024 23:44:38 +0000
Subject: [PATCH 1/2] fix: correct zero row count display in DataFrame from
 table view

---
 bigframes/core/nodes.py              | 4 +++-
 tests/system/small/test_dataframe.py | 7 +++++++
 2 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/bigframes/core/nodes.py b/bigframes/core/nodes.py
index e65040686e..9c30bc762b 100644
--- a/bigframes/core/nodes.py
+++ b/bigframes/core/nodes.py
@@ -508,6 +508,7 @@ class GbqTable:
     table_id: str = field()
     physical_schema: Tuple[bq.SchemaField, ...] = field()
     n_rows: int = field()
+    is_view: bool = field()
     cluster_cols: typing.Optional[Tuple[str, ...]]
 
     @staticmethod
@@ -523,6 +524,7 @@ def from_table(table: bq.Table, columns: Sequence[str] = ()) -> GbqTable:
             table_id=table.table_id,
             physical_schema=schema,
             n_rows=table.num_rows,
+            is_view=(table.table_type == "VIEW"),
             cluster_cols=None
             if table.clustering_fields is None
             else tuple(table.clustering_fields),
@@ -603,7 +605,7 @@ def variables_introduced(self) -> int:
 
     @property
     def row_count(self) -> typing.Optional[int]:
-        if self.source.sql_predicate is None:
+        if self.source.sql_predicate is None and not self.source.table.is_view:
             return self.source.table.n_rows
         return None
 
diff --git a/tests/system/small/test_dataframe.py b/tests/system/small/test_dataframe.py
index cfd6efe9bd..6793a2bb6c 100644
--- a/tests/system/small/test_dataframe.py
+++ b/tests/system/small/test_dataframe.py
@@ -1521,6 +1521,13 @@ def test_shape(scalars_dfs):
     assert bf_result == pd_result
 
 
+def test_view_shape(session):
+    view_df = session.read_gbq("bigframes-dev.bigframes_tests_sys.base_table_view")
+    table_ref = session.read_gbq("bigframes-dev.bigframes_tests_sys.base_table")
+
+    assert view_df.shape == table_ref.shape
+
+
 def test_len(scalars_dfs):
     scalars_df, scalars_pandas_df = scalars_dfs
     bf_result = len(scalars_df)

From 24cd146ccacab6f203e931e4f48d784e5e75b951 Mon Sep 17 00:00:00 2001
From: Huan Chen <huanc@google.com>
Date: Tue, 8 Oct 2024 00:23:09 +0000
Subject: [PATCH 2/2] update logic and test

---
 bigframes/core/nodes.py              |  6 +++---
 tests/system/small/test_dataframe.py | 21 +++++++++++++++++----
 2 files changed, 20 insertions(+), 7 deletions(-)

diff --git a/bigframes/core/nodes.py b/bigframes/core/nodes.py
index 9c30bc762b..d750ee63fb 100644
--- a/bigframes/core/nodes.py
+++ b/bigframes/core/nodes.py
@@ -508,7 +508,7 @@ class GbqTable:
     table_id: str = field()
     physical_schema: Tuple[bq.SchemaField, ...] = field()
     n_rows: int = field()
-    is_view: bool = field()
+    is_physical_table: bool = field()
     cluster_cols: typing.Optional[Tuple[str, ...]]
 
     @staticmethod
@@ -524,7 +524,7 @@ def from_table(table: bq.Table, columns: Sequence[str] = ()) -> GbqTable:
             table_id=table.table_id,
             physical_schema=schema,
             n_rows=table.num_rows,
-            is_view=(table.table_type == "VIEW"),
+            is_physical_table=(table.table_type == "TABLE"),
             cluster_cols=None
             if table.clustering_fields is None
             else tuple(table.clustering_fields),
@@ -605,7 +605,7 @@ def variables_introduced(self) -> int:
 
     @property
     def row_count(self) -> typing.Optional[int]:
-        if self.source.sql_predicate is None and not self.source.table.is_view:
+        if self.source.sql_predicate is None and self.source.table.is_physical_table:
             return self.source.table.n_rows
         return None
 
diff --git a/tests/system/small/test_dataframe.py b/tests/system/small/test_dataframe.py
index 6793a2bb6c..6ee9fb8247 100644
--- a/tests/system/small/test_dataframe.py
+++ b/tests/system/small/test_dataframe.py
@@ -1521,11 +1521,24 @@ def test_shape(scalars_dfs):
     assert bf_result == pd_result
 
 
-def test_view_shape(session):
-    view_df = session.read_gbq("bigframes-dev.bigframes_tests_sys.base_table_view")
-    table_ref = session.read_gbq("bigframes-dev.bigframes_tests_sys.base_table")
+@pytest.mark.parametrize(
+    "reference_table, test_table",
+    [
+        (
+            "bigframes-dev.bigframes_tests_sys.base_table",
+            "bigframes-dev.bigframes_tests_sys.base_table_view",
+        ),
+        (
+            "bigframes-dev.bigframes_tests_sys.csv_native_table",
+            "bigframes-dev.bigframes_tests_sys.csv_external_table",
+        ),
+    ],
+)
+def test_view_and_external_table_shape(session, reference_table, test_table):
+    reference_df = session.read_gbq(reference_table)
+    test_df = session.read_gbq(test_table)
 
-    assert view_df.shape == table_ref.shape
+    assert test_df.shape == reference_df.shape
 
 
 def test_len(scalars_dfs):